In the recent drive for higher integration and operating speeds in LSI devices, it is desired to miniaturize the pattern rule. Great efforts have been devoted for the development of the micropatterning technology using deep-ultraviolet (deep-UV) or vacuum-ultraviolet (VUV) lithography. The photolithography using KrF excimer laser (wavelength 248 nm) as the light source has already established the main role in the commercial manufacture of semiconductor devices. The lithography using ArP excimer laser (wavelength 193 nm) is under investigation to enable further miniaturization and has reached the stage of prototype manufacture experiments. However, the ArF excimer laser lithography has not matured so that many problems must be overcome before the technology can be applied to an industrial scale of semiconductor manufacture.
The requisite properties for the resist materials complying with the ArF excimer laser lithography include transparency at wavelength 193 nm and dry etch resistance. Resist materials comprising as a base resin poly(meth)acrylic acid derivatives having bulky acid-labile protective groups as typified by 2-ethyl-2-adamantyl and 2-methyl-2-adamantyl groups were proposed as having both the properties (JP-A 9-73173 and JP-A 9-90637). Since then, a variety of materials have been proposed. Most of them commonly use resins having a highly transparent backbone and a carboxylic acid moiety protected with a bulky tertiary alkyl group.
While these prior art resist materials suffer from several problems, the most serious problem is the unevenness of a fine line size generally known as “line edge roughness” (LER). There is a strong desire to solve the problem since such unevenness largely affects the performance of semiconductor devices manufactured. If it suffices to provide the pattern with a smooth finish, the object may be attained to some extent by setting the molecular weight of a resin at a relatively low level or by selecting such a photoacid generator that the acid resulting therefrom is more mobile. These approaches, however, extremely exacerbate such properties as exposure dose dependency, pattern density dependency, and mask fidelity, and do not entail a reduction of LER because fine fluctuations of the mask are enlarged so that the line size itself rather becomes uneven. While a further reduction of the pattern rule is invariably required, there is a demand for a resist composition which is essentially improved in LER without the sacrifice of resolution while exerting excellent performance in terms of sensitivity, substrate adhesion and etch resistance.
Studies have also been made on photoacid generators. In prior art chemically amplified resist compositions for lithography using KrF excimer laser, photoacid generators capable of generating alkane- or arene-sulfonic acid are used. However, the use of these photoacid generators in chemically amplified resist compositions for ArF lithography results in an insufficient acid strength to scissor acid labile groups on the resin, a failure of resolution or a low sensitivity. Thus these photoacid generators are not suited for the fabrication of microelectronic devices.
For the above reason, photoacid generators capable of generating perfluoroalkanesulfonic acid having a high acid strength are generally used in ArF chemically amplified resist compositions. Perfluorooctanesulfonic acid and derivatives thereof (collectively referred to as PFOS) are considered problematic with respect to their stability (or non-degradability) due to C—F bonds, and biological concentration and accumulation due to hydrophobic and lipophilic natures. With respect to perfluoroalkanesulfonic acids of 5 or more carbon atoms and derivatives thereof, the same problems are pointed out.
Facing the PFOS-related problems, manufacturers made efforts to develop partially fluorinated alkane sulfonic acids having a reduced degree of fluorine substitution. For instance, JP-A 2004-531749 describes the development of α,α-difluoroalkanesulfonic acid salts from α,α-difluoroalkene and a sulfur compound and discloses a resist composition comprising a photoacid generator which generates such sulfonic acid upon irradiation, specifically di(4-tert-butylphenyl)iodonium 1,1-difluoro-2-(1-naphthyl)ethanesulfonate. JP-A 2004-2252 describes the development of α,α,β,β-tetrafluoroalkanesulfonic acid salts from α,α,β,β-tetrafluoro-α-iodoalkane and sulfur compound and discloses a photoacid generator capable of generating such a sulfonic acid and a resist composition comprising the same. JP-A 2002-214774 discloses photoacid generators having difluorosulfoacetic acid alkyl esters (e.g., 1-(alkoxycarbonyl)-1,1-difluoromethanesulfonate) and difluorosulfoacetic acid amides (e.g., 1-carbamoyl-1,1-difluoromethanesulfonate) although their synthesis method is lacking. Furthermore, JP-A 2004-4561 discloses triphenylsulfonium (adamantan-1-ylmethyl)oxycarbonyldifluoromethanesulfonate although its synthesis method is lacking; JP-A 2006-306856 discloses triphenylsulfonium alkyloxycarbonyldifluoromethanesulfonates having a lactone structure and analogs; and JP-A 2007-145797 discloses triphenylsulfonium 2-acyloxy-1,1,3,3,3-hexafluoropropanesulfonate and analogs.
As far as the inventors empirically confirmed, undesirably these compounds suffer from problems including difficult compound design due to limited starting reactants (JP-A 2004-531749, JP-A 2004-2252), low solubility (JP-A 2002-214774, JP-A 2004-4561, JP-A 2006-306856), and increased hydrophobicity due to many fluorine atoms (JP-A 2007-145797).
With respect to the immersion lithography, there remain some other problems. Minute water droplets are left on the resist and wafer after the immersion exposure, which can often cause damages and defects to the resist pattern profile. The resist pattern after development can collapse or deform into a T-top profile. There exists a need for a patterning process which can form a satisfactory resist pattern after development according to the immersion lithography.